Topic on User talk:JakobVoss

Hello Jakob, thanks for creating the mix'n'match-list of author-related notations in RVK to Wikidata Concepts. I am a bit confused, about RVK and whether it's really a 1:1 mapping. Many authors have in RVK at least two entries ("works", "literature about" - e.g. Kant) or three entries ("collection", "work", "lit. about" - e.g. Mortier, Erwin)

If I understand RVK, the first notation has no hierarchical aspect; Maybe we should match all RVK entries to the corresponding author and add a qualifier e.g. of (P642) to edited volume (Q1711593), work (Q386724) or secondary literature (Q18891569)? Thank you, --Mfchris84 (talk) 05:36, 11 September 2018 (UTC)

RVK is a hierarchy so only the top most matching RVK classes matter. For example the RVK notation for Kant is "BF 4045 - BF 4046". Just ignore the entries below! Independent from that some famous authors occur at multiple places, e.g. Kant also has "CF 5000 - CF 5017". Apart from that just think of RVK notations as identifiers. I don't understand your question about qualifiers and see no need to use other qualifiers but mapping relation type (P4390) for mapping topics (not authors).

Form a librarians view its in the beginning a bit confusing, to think notations as external identifiers. I see you have already removed RVK claims in bibliographic items on Wikidata. Maybe it will be good to have clarification on wikidata about the usage of library classification systems. Dewey Decimal Classification (P1036) or Library of Congress Classification (P1149) are still used as classification of bibliographic items.

Yes, DDC and LCC may be changed later but I did not want to start a revolution. By now we have Regensburg Classification (P1150) and Basisklassifikation (P5748) used as identifiers and the other classification properties uncontrolled. It will be easier to migrate all classification from use for bibliographic items to use for mapping each classification with Wikidata items when there will be more mappings available.